04_time_group_event_analysis

library(plotly)
ggplotly(
  
  ggplot(filter(pronoun_use_time_group_collective, num_comments > 1500), 
         aes(x=time_block,y=collective,name=subreddit,color=all_topics, size = num_comments)) +
    geom_point() + geom_smooth() +  geom_vline(xintercept  = 316)
  
)

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
Computation failed in `stat_smooth()`:
x has insufficient unique values to support 10 knots: reduce k.

pronoun_use_time_group_collective %>%
  mutate(day = round(time_block)) %>%
  group_by(day,topic_level2) %>%
  summarise(
    num_comments_day = sum(num_comments),
    collective = sum(collective*num_comments)/num_comments_day
  ) %>% #filter(subreddit == "The_Donald") %>% 
  ggplot(aes(x=day,y=collective))+#,size = num_comments_day, color = topic_level2)) +
  geom_point() + geom_smooth() + geom_vline(xintercept  = 316) + facet_wrap(~topic_level2) +
  scale_y_continuous(limits = c(0.05,0.3))

NA

Create offsets to normalize use rates

pronoun_use_time_group_collective <- pronoun_use_time_group_collective 
ggplotly(
  ggplot(filter(pronoun_use_time_group_collective, subreddit %in% c("The_Donald","hillaryclinton")) ,
         aes(x=time_block,y=ZNcollective, color = subreddit))+#,size = num_comments_day, color = topic_level2)) +
  geom_point() + geom_smooth() + geom_vline(xintercept  = 316) +
  scale_y_continuous(limits = c(-50,200))
)

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
Removed 45 rows containing non-finite values (stat_smooth).